screenshot-reader

name: screenshot-reader description: Read screenshots/images and extract text + actionable insights. homepage: https://docs.openclaw.ai metadata: { "openclaw": { "emoji": "🖼️", "requires": {"inputs": ["image"]} }

}

screenshot-reader

Use this skill when Ryan drops a screenshot (X posts, dashboards, receipts, UI screens) and wants the content read + turned into actions.

What it does

Given an image, return:

1. Exact extracted text (as faithfully as possible)
2. Summary (1-3 sentences)
3. Ideas / next actions (bullet list)
4. If it’s a social post: suggested reply + suggested quote-tweet

How to use (Ryan)

1. Upload a screenshot in Telegram.
2. Say what you want: "summarize", "pull ideas", "draft reply", "extract handles", etc.

How I run it (PD)

• Use built-in vision to read the screenshot.
• If handles/links are present: normalize them to copy/paste.
• If it’s a UI flow: identify the next 3 clicks.

Output format (default)

TEXT:

SUMMARY:

IDEAS / ACTIONS:

• ...

DRAFT REPLY (optional): ...

DRAFT POST (optional): ...